Medicines are given for cancer patients Clinical trials can help save or extend lives.
But despite thousands of studies in the United States each year, only 3 to 5 percent of eligible patients participate in studies of new treatments.
Triomics, a generative AI startup, claims it can significantly reduce the time it takes doctors to match patients with trials.
Doctors’ recommendations are often the key to accepting patients. However, busy oncologists and nurses often lack the time to learn about all the clinical trials that may be appropriate for their patients.
I am not a doctor and therefore do not know the daily challenges of medical staff in oncology. But unfortunately, I know from personal experience how difficult it is to find clinical trials for cancer patients. I spent countless hours browsing Clinicaltrials.gov, a website and database that lists thousands of ongoing trials, when my father was ill. And just in March, I spent half a Saturday trying to find a clinical trial for a friend who has stage IV cancer. Her doctor only offered one trial, so she asked me if there were other options.
Because most clinical trials have complex criteria, dozens of factors such as cancer stage, mutations and previous treatments often determine eligibility. Medical staff often spend hours manually reviewing a patient’s medical record and finding an appropriate clinical trial. However, due to the shortage of professionals in oncology, many cancer patients are not offered participation or miss their participation window.
The company was founded by former MIT biotech researcher Sarim Khan and Adobe AI scientist Hrituraj Singh. The couple, who have been friends since college, decided to start Triomics in 2021 after realizing that advances in generative AI and LLMs could help extract data from electronic health records (EHR) to inform appropriate clinical trials for Find cancer patients in minutes instead of hours.
Khan and Singh joined Y Combinator in winter 2021, working on an LLM designed specifically for cancer centers and oncology departments in hospital systems.
Three years later, Triomics announces that six cancer centers and hospitals are actively using or testing the LLM, and the company plans to double that number by the end of the year. And now the company has raised $15 million in Series A funding from Lightspeed, Nexus Venture Partners, General Catalyst and Y Combinator to help it further develop its platform and launch it to new customers.
While reducing the time it takes to get patients into clinical trials may seem like the most immediately valuable application of Triomics software, Khan says Triomics is much more than a clinical trial company. “Doctors use it for different use cases that I could just keep talking about,” he said.
After Triomics’ LLM, which the company calls OncoLLM, “reads” the patient’s medical record, the data could be used to prepare doctors and other medical staff for patient visits or to submit cancer data with details about the organs affected and the stage of progression state regulatory authorities.
Of course, Triomics is not the only one tackling this area. Other startups performing AI clinical trial matching include Deep 6 AI, QuantHealth and Trajectory
However, Khan believes Triomics is one of the few startups processing large volumes of datasets specifically for cancer centers.